Personalization and Failure Testing of Dual Switch Gene Drives in Lung Cancer

Project Summary Different patients with Non-small-cell lung cancers (NSCLC) can harbor mutations that result in constitutively activated versions of tyrosine kinases (e.g. EGFR, RET, ALK, ROS1, TRK) that can be precisely targeted with inhibitors. However, tyrosine kinase inhibitors are vulnerable to existing, known and unknown, drug resistance mechanisms found in tumors. This results in a game of molecular “whack-a-mole” whereby, resistance evolution appears, the mechanism is isolated, drugs are administered to combat that drug resistance, and then resistance re-emerges until no effective therapies remain. This process of reverse engineering drug resistance has been a losing battle with a high cost for patients. A promising approach to combat the challenge of resistance evolution is to design and test cell therapies that can sense the therapeutic environment and respond through synthetic biology circuits to reproducibly control evolutionary trajectories. We propose a synthetic biological technology with proof-of-concept function in mammalian cells that we term “dual-switch selection drives”. These drives use inducible drug resistance to create a cell therapy that can engineer a tumor’s evolution in situ. The first switch senses the presence of a dimerizer molecule to create reversible drug resistance. Using the mathematical rules of biophysics and evolution, our cell therapy calculates a response to small molecules and produces a tunable amount of cellular fitness that competes with pre-existing drug resistance variants in a tumor. A second switch with a suicide gene payload hitchhikes on this evolution guided cell therapy until the selection drive cells comprise the majority of the tumor. Then, at the flip of a second switch, a locally diffusible toxin is produced that kills all cells--gene drive or pre-existing resistance mutants of any molecular origin--through a bystander effect. This technology works with the existing standard of care drugs in NSCLC to produce localized combination therapy that can eradicate pre-existing resistance regardless of the molecular mechanism. Therefore, instead of responding to and combatting evolution, we use forward engineering of cell therapies to direct evolution. In Aim 1 we will use nonintuitive insights from stochastic models of the evolutionary stability of our designs to propose further optimized selection drives. Aim 2 expands our forward engineering approach by pushing our model driven design of safety and efficacy towards the spatial, cellular, and microenvironmental heterogeneity present in NSCLC. Aim 3 proposes to move evolutionary proof-of-concept experiments into primary human organoids from NSCLC patients with activating mutations in EGFR. Beyond practical testing of a technology, we will also “build to understand” the basic cancer biology of resistance evolution.